Analytical Approaches To World Music 2.1 (2012) 87-137 CantoCore: A New Cross-Cultural Song Classification Scheme Patrick E. Savage, Emily Merritt, Tom Rzeszutek, and Steven Brown Department of Psychology, Neuroscience & Behaviour, McMaster University Classification of organisms and languages has long provided the foundation for studying biological and cultural history, but there is still no accepted scheme for classifying songs cross-culturally. The best candidate, Lomax and Grauer’s “Cantometrics” coding scheme, did not spawn a large following due, in part, to concerns about its reliability. We present here a new classification scheme, called “CantoCore”, that is inspired by Cantometrics but that emphasizes its “core” structural characters rather than the more subjective characters of performance style. Using both schemes to classify the 30 songs from the Cantometrics Consensus Tape, we found that CantoCore appeared to be approximately 80% more reliable than Cantometrics. Nevertheless, Cantometrics still demonstrated significant reliability for all but its instrumental characters. Future multidisciplinary applications of CantoCore and Cantometrics to the cross-cultural study of musical similarity, musical evolution, musical universals, and the relationship between music and culture will provide the true test of each scheme’s value. Keywords: music, song, classification, CantoCore, Cantometrics, reliability, musilinguistic continuum This is a single-spaced version of the article. The official version with page numbers 87-137 can be viewed at http://aawmjournal.com/articles/2012a/Savage_Merritt_Rzeszutek_Brown_AAWM_Vol_2_1.pdf INTRODUCTION according to geographic and ethnolinguistic usical classification is a topic that classifications do not use an explicitly musical has received scant attention since classificatory framework. the heyday of comparative A consideration of the historical roots of the musicology during the first half of field shows that classification was central to the first the 20th century. Fields like definition of comparative musicology: biologyM and linguistics have long relied on [C]omparative musicology has as its task the classification as the starting point for developing comparison of the musical works— broader theories, such as Darwin’s (1859) theory of especially the folksongs—of the various evolution and Jones’ (1807) theory of prehistoric peoples of the earth for ethnographical connections among speakers of Indo-European purposes, and the classification of them languages. Today, global linguistic classification according to their various forms (Adler databases such as the Ethnologue (Lewis 2009) and 1885, 14). the World Atlas of Language Structures Although classification, comparison, and (Haspelmath et al. 2005) are fundamental to the ethnography were all equal parts of this original study of language evolution, linguistic universals, definition, the field later changed its name to and human history (Currie and Mace 2009; Dunn et “ethnomusicology” and developed a methodological al. 2011; Atkinson 2011). Musicology, in contrast, emphasis on single-culture ethnography over cross- never entered into a comfortable relationship with cultural classification and comparison. This was cross-cultural classification, despite early attempts part of a broader trend in anthropology in the wake in that direction (Hornbostel and Sachs 1914; of World War II toward cultural relativism and Lomax 1968). Even global music collections like away from universalism (Geertz 1973). One the Garland Encyclopedia of World Music (Stone et outcome of this shift was the recognition of a al. 1998) and Smithsonian Global Sound theoretical distinction between “etic” (objective, (http://glmu.alexanderstreet.com) that are organized outsider) and “emic” (subjective, insider) theories of 1 CantoCore: A New Cross-Cultural Song Classification Scheme classification (Harris 1976). This dichotomy nicely based on multidimensional, musilinguistic spectra characterizes the paradigmatic difference between could be helpful in fields as diverse as early comparative musicology and contemporary ethnomusicology, neuroscience, and evolutionary ethnomusicology. Ethnomusicologists have largely biology for understanding connections between rejected etic and/or acoustic classification schemes, music and language (Darwin 1871; Feld and Fox despite pleas for pluralism in approaches to world 1994; Wallin, Merker, and Brown 2000; Patel musics (Merriam 1982; Nettl 2005; Agawu 2010). 2008). Although the goal of classifying musics acoustically Multi-dimensional, musilinguistic spectra are in presents many challenges—for example, the need fact a major design feature of the best-established that classification schemes be universally song-classification scheme to date, “Cantometrics” applicable—these challenges do not a priori (Lomax and Grauer 1968; Lomax 1976). invalidate cross-cultural classification (but see Hood Cantometrics classifies songs according to 37 1971; Blacking 1973; McLeod 1974). acoustic characters related to their structure, Along these lines, there are two major performance style, and instrumental methodological challenges to classifying music accompaniment. Each character contains between 3 cross-culturally. One challenge is specific to and 13 character-states, which are ordered along a instrumental music: how do we ensure that we are social continuum from “individualized” to comparing like with like when different cultures use “groupy.” This continuum can be thought of equally different instruments with differing acoustic well as a musilinguistic continuum, since speech features, production mechanisms, and tuning tends to be more individual-oriented and song more systems (Ellis 1885)? The second is specific to group-oriented. vocal music: how can we design a classification Applying this scheme to a global sample of system that is broad enough to accommodate all thousands of songs from hundreds of cultures, musical cultures while maintaining a distinction Lomax found that global song diversity was between “song” and “speech”? While the organized into 10 major stylistic families that also instrumental classification scheme of Hornbostel correlated with extra-musical features of social and Sachs (1914) is still widely used today, there structure and historical contact. Critics generally remains no widely accepted song-classification applauded this ground-breaking attempt to scheme. quantitatively address the relationship between One solution to the problem of song music and culture and supported its broad findings, classification is to see the relationship between despite some concerns over methodological issues music and language as a continuum—a regarding sampling, treatment of intra-cultural “musilinguistic” spectrum (Brown 2000)—rather diversity, and the interpretation of correlations than as a contrast between two discrete domains. A between music and social structure (Naroll 1969; truly universal approach cannot exclude “non- Driver 1970; Downey 1970; Nettl 1970; Maranda musical” vocalizations but must accommodate any 1970; Henry 1976; Erickson 1976; Dowling and type of vocalization sitting along the musilinguistic Harwood 1986; Grauer 2005; Leroi and Swire spectrum of communicative forms from speech, to 2006). However, many critics were divided over songs, to everything in between. While Sachs Lomax’s emphasis on performance style over song (1943) proposed such a spectrum in his distinction structure. Lomax’s agenda in creating Cantometrics between “logogenic” (word-born) and “melogenic” was to replace Western musicology’s traditional (melody-born) songs, there is a need for a emphasis on musical structure and notation—which classification scheme that can accommodate the he and many others saw as being Eurocentric and diversity of ways in which song-features can elitist (Lomax 1959; Feld and Fox 1994)—with a independently vary across multiple musilinguistic more performance-oriented system. While some spectra. For example, some songs can have irregular critics supported the development of measurements “speech-like” (parlando) rhythms but use discrete of performance characters such as “nasality” and “music-like” pitches, while others can have metric “rasp,” others were concerned that such characters “music-like” rhythms but use indeterminate were overly subjective and thus unreliable (Downey “speech-like” pitches. A classificatory approach 1970; Maranda 1970). 2 Analytical Approaches To World Music 2.1 (2012) 87-137 The principal objective of the current study is to structural characters not present in Cantometrics, present a detailed analysis of a new universal song- most notably those related to scales and rhythms. classification scheme. We call it “CantoCore” Finally, the scheme is designed to accommodate because of its emphasis on the “core” structural musical forms at all points along the musilinguistic characters of song. The scheme takes its lead from spectrum, from a simple sentence to the most the updated 1976 version of Cantometrics but complexly-textured responsorial polyphony. The focuses only on characters of song-structure rather current study also includes a test of the inter-rater than performance-style or instrumentation (see reliability of song codings, comparing 1) CantoCore Figure 1), because of our prediction that structural vs. Cantometrics, and 2) the structural characters of characters should be more reliable. We have Cantometrics vs. its performance and instrumental reorganized, supplemented, and attempted